Testing the replication hypothesis: When the data set is subject to gross error

Between the time OLS estimates are computed on N observations and replication is attempted, the data matrix accumulates C rows of gross errors. Fitting the reported OLS estimates to the contaminated data will produce N-C residuals which agree exactly with the original residuals. This fact serves as...

Full description

Saved in:
Bibliographic Details
Published inEconomics letters Vol. 34; no. 1; pp. 49 - 53
Main Authors Levy, David M., Feigenbaum, Susan
Format Journal Article
LanguageEnglish
Published Amsterdam Elsevier B.V 1990
Elsevier
North Holland
SeriesEconomics Letters
Subjects
Econometric models
Finland
Schätztheorie
Theorie
Online AccessGet full text

Cover

More Information
Summary:Between the time OLS estimates are computed on N observations and replication is attempted, the data matrix accumulates C rows of gross errors. Fitting the reported OLS estimates to the contaminated data will produce N-C residuals which agree exactly with the original residuals. This fact serves as the basis of a test for replication.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
content type line 23
ISSN:0165-1765
1873-7374
DOI:10.1016/0165-1765(90)90180-9